P.J. Lea, Department of Biological Sciences, Lancaster University, Lancaster LA1 4YQ, UK. Email: email@example.com
Interest in plant asparagine has rapidly taken off over the past 5 years following the report that acrylamide, a neurotoxin and potential carcinogen, is present in cooked foods, particularly carbohydrate-rich foods such as wheat and potatoes which are subjected to roasting, baking or frying at high temperatures. Subsequent studies showed that acrylamide could be formed in foods by the thermal degradation of free asparagine in the presence of sugars in the Maillard reaction. In this article, our current knowledge of asparagine in plants and in particular its occurrence in cereal seeds and potatoes is reviewed and discussed in relation to acrylamide formation. There is now clear evidence that soluble asparagine accumulates in most if not all plant organs during periods of low rates of protein synthesis and a plentiful supply of reduced nitrogen. The accumulation of asparagine occurs during normal physiological processes such as seed germination and nitrogen transport. However, in addition, stress-induced asparagine accumulation can be caused by mineral deficiencies, drought, salt, toxic metals and pathogen attack. The properties and gene regulation of the enzymes involved in asparagine synthesis and breakdown in plants are discussed in detail.
Asparagine was the first amino acid to be isolated from plants, 200 years ago (Vauquelin & Robiquet, 1806), after the characteristic cubic crystals had been observed in the concentrated solutions of the sap of Asparagus species by Delaville (1802). Asparagine is an amino acid amide that has a molecular mass of 132.12 and an isoelectric point of 5.41. Although soluble in both acids and alkalis, asparagine is only moderately soluble in water and readily forms white monohydrate crystals. Furthermore, asparagine has a N:C ratio of 2:4 (Fig. 1), which makes it an efficient molecule for the storage and transport of nitrogen in living organisms.
Although it was not initially thought to be a constituent of proteins, asparagine was shown to be present in the enzymic digest of the Brazil nut seed storage protein edestin (Damodaran, 1932) and is now known to be present in most proteins. Although it lacks the negative charge of the free carboxylic acid group of aspartate, asparagine retains some polarity and frequently plays a key role in the active site of enzymes (Mansfield et al., 2006). In addition, a range of oligosaccharides may be attached to the amide group of protein-bound asparagine, catalysed by glycosyltransferases (Lerouge et al., 1998; Bencur et al., 2005).
The interest in plant asparagine has really taken off over the past 5 years following the report that the neurotoxin and potential carcinogen acrylamide was present in cooked foods, particularly carbohydrate-rich foods such as wheat and potatoes which are subjected to roasting, baking or frying at high temperatures (Tareke et al., 2002). Subsequent studies showed that acrylamide could be formed in foods by the thermal degradation of free asparagine in the presence of sugars in the Maillard reaction (Mottram et al., 2002; Stadler et al., 2002; Zyzak et al., 2003), leading to a series of studies (described later in the review) on the levels of asparagine in cereals and potatoes and their relationship to the levels of acrylamide in baked and fried products. It is therefore appropriate to review our current knowledge of asparagine in plants, including its distribution, synthesis and degradation, biological role and relationship to acrylamide formation during processing.
The occurrence and accumulation of asparagine in plants
Asparagine has a high ratio of N:C, is a substrate for only a few enzymatic reactions in the soluble form and has little net charge under physiological conditions. It therefore forms an ideal storage compound and accumulates under a range of situations. Similarly, it is a major transported form of nitrogen, particularly in leguminous plants.
Transport of reduced nitrogen in annuals
Asparagine is the major transport compound in the xylem from the root to the leaves and in the phloem from the leaves to the developing seeds in a range of plants. A considerable amount of information has been obtained on these transport processes by Pate, Atkins and colleagues, following their elegant and detailed studies on legumes (Pate, 1980). In nitrogen-fixing Lupinus albus, asparagine was the major amino acid in all plant parts and could account for 60–80% of the total amino acids in nodulated roots, leaves and fruits (Pate et al., 1981). Asparagine was also the major component of the xylem (45–50%) and less so of the phloem (20–30%), where glutamine also made a significant contribution. Interestingly, the proportions remained relatively constant during the 60-day growing period (Pate et al., 1979; Atkins et al., 1983). However, if the roots were treated with argon, thus removing nitrogen and preventing fixation, there were rapid decreases in the concentrations of amino acids and in the proportion of asparagine, in both the xylem and phloem (Pate et al., 1984).
As we have seen above, asparagine is the major end product of nitrogen fixation in Lupinus species and is present in high concentrations in the nodules and root xylem exudates. This is also the case in other temperate legumes such as Medicago (Snapp & Vance, 1986; Ta et al., 1986) and Pisum (Fig. 2) (Scharff et al., 2003). However, the precise pathway by which the ammonia is incorporated into asparagine within the nodule is still a matter of debate (Prell & Poole, 2006). Even in peanut (Arachis hypogea), which is known to contain the very unusual amide γ-methyleneglutamine (Fowden, 1954), asparagine was the major amino acid in the nodules and xylem exudates of nitrogen-fixing plants (Peoples et al., 1986). However, not all nitrogen-fixing legumes use asparagine as the major nitrogen transport compound. Tropical legumes belonging to the Phaseolae tribe, e.g. soybean (Glycine max) (Rainbird et al., 1984; Herridge & Peoples, 1990) and cowpea (Vigna unguiculata) (Peoples et al., 1985; Atkins et al., 1988), use the ureides allantoin and allantoic acid, derived from purine metabolism (Smith & Atkins, 2002).
Recent data published by do Amarante et al. (2006) have proved useful in improving our understanding as to when legumes use asparagine rather than ureides as nitrogen transport compounds in the xylem. The data shown in Fig. 3 clearly indicate that ureides are the major form of N transport compounds in nodulated G.max, V.unguiculata and Phaseolus aureus, whereas amino acids and nitrate predominate when the plants are not fixing nitrogen. In the species that form small or zero amounts of ureides (Crotalaria juncea, Pisum sativum and L.albus), only amino acids were found in the xylem sap of nodulated plants, whereas both nitrate and amino acids were found in the plants not undergoing nitrogen fixation.
A more detailed analysis of the xylem amino acid fractions of six different plants (do Amarante et al., 2006) indicated that under most circumstances, asparagine was still the major form of N transport, although glutamine was also prominent in the nitrogen-fixing, symbiotic, ureide-forming plants (Fig. 4). The difference in asparagine between symbiotic and non-symbiotic plants was much less apparent in the non-ureide-forming plants. The data clearly indicate that even in those plants classified as ureide formers, a significant quantity of nitrogen was also exported in the xylem as the amides, glutamine and asparagine. The importance of the proportion of asparagine in the xylem as an indicator of the nitrogen status of a plant can be seen from the data shown in Fig. 5. Four days of maintaining the plants in nitrogen-free conditions caused a reduction in the proportion of asparagine and an increase in aspartate in all the plants tested, irrespective of whether nitrate or nitrogen fixation was the original source of nitrogen (do Amarante et al., 2006).
In nodulated soybean, such changes in asparagine transport in the xylem were associated with a decrease in asparagine synthetase (AS) activity of the nodules (Lima & Sodek, 2003).
Transport and storage of reduced nitrogen in perennials
Deciduous trees and other woody plants store reduced nitrogen in the form of either vegetative storage proteins or amino acids over the winter period. The nitrogen is rapidly remobilised in the spring and transported to the developing leaf and flower buds in the form of soluble amino acids (Millard, 1996). In addition, there is a massive recycling of nitrogen following the metabolism of phenylalanine and tyrosine to form lignin (Cantón et al., 2005). Early work on apple (Malus domesticus) indicated that asparagine and arginine were the major transport compounds in the phloem, with the former used for short distance transport and the latter for long distance transport (Tromp & Ovaa, 1971). Data obtained by Malaguti et al. (2001) showed that in apple, asparagine accounted for more than half of the total amino acid nitrogen in the xylem, followed by arginine, glutamine and aspartate, at the time of full bloom. Later in the season, asparagine still predominated in the xylem but was derived from remobilisation and also nitrogen taken up by the roots (Tromp & Ovaa, 1976).
Asparagine often in combination with glutamine and arginine has been shown to be the major nitrogen transport compound in citrus (Citrus unshiu) (Kato, 1981), poplar (Populus spp.) (Escher et al., 2004), beech (Fagus sylvatica) (Nahm et al., 2006) and cherry (Prunus avium) (Millard et al., 2006). Aidar et al. (2003) studied a Brazilian Atlantic forest community, aiming to characterise the strategies involved in tree N acquisition and transport during different phases of forest succession. For the pioneer species, asparagine was the main nitrogen transport compound, comprising nearly 50% of the xylem nitrogen. The leguminous early secondary species transported a variety of compounds, but in all tree species, asparagine accounted for 24–76% of the xylem nitrogen.
In contrast, in evergreen conifers, there is evidence that asparagine does not play a major role in the storage and transport of nitrogen in the mature tree. In Scots pine (Pinus sylvestris), arginine and glutamine were the major forms of soluble nitrogen in the bark, wood and needles and were increased by N fertilisation (Nordin et al., 2001). Significant quantities of arginine were lost when senescent needles were shed, particularly from well-fertilised trees (Näsholm, 1994). High concentrations of arginine were also detected in the needles of Norway spruce (Picea abies), mostly when minerals other than nitrogen were limiting (Ericsson et al., 1993). Schneider et al. (1996) carried out a direct comparison of the conifer P.abies and deciduous Fagus sylvatica grown under field conditions of high nitrogen. The amino acid content of the leaves, phloem and xylem of the two trees were determined during the period from April to September. In P.abies, soluble asparagine was not detected in any appreciable amount at any time in any tissue throughout the growing season. In F.sylvatica, asparagine was the predominant amino acid in the leaves and xylem but not phloem in April and May only. However, glutamine and arginine were often present in high concentrations in both trees.
It is clear, therefore, that asparagine can act as a storage and transport molecule, often at the same time within the plant. However, arginine and glutamine are also able to carry out the same role. There is also evidence of species differences in the use of these compounds, as well as temporal and tissue differences.
King & O’Donoghue (1995) have examined in some detail the metabolic processes taking place in harvested spears (shoot tips) of asparagus (Asparagus officinalis), the original source of asparagine. Initially, the major soluble amino acid in the tips was glutamine, but following harvest, the asparagine concentration increased from 6.5 mg g−1 dry weight (DW) to 28.6 mg g−1 in 48 h. Within the 48-h storage period, the rate of respiration fell dramatically, and there was an almost complete depletion of soluble monosaccharides. During the second 24-h period, the concentration of ammonium also increased (Eason et al., 1996). It has been proposed that the asparagine content could be used as a direct ’freshness test’ of asparagus spears (Hurst et al., 1998), as increases from 50 to more than 400 μmol g−1 DW (Fig. 6) were detected. Large increases in asparagine and glutamine concentrations and a reduction in soluble sugars were also detected in broccoli florets for up to 80 h following harvest (Downs et al., 1997). The data suggested that following senescence induced by the harvesting of either asparagus spears or broccoli florets, asparagine acts as a store of nitrogen, when the supply of soluble carbohydrate is severely reduced.
Asparagine has also been shown to be a major product during metabolism in senescing leaves, although it should be noted that the majority of experiments were carried out on leaves that had been detached from the plant and did not senesce naturally. Mothes (1926, 1940) showed that when leaves of Vicia faba and Phaseolus multiflorus were detached, the protein level decreased with an equivalent rise in soluble amino acids, in particular glutamine; this was followed by a rapid increase in asparagine. In detached leaves of tobacco, work by Vickery et al. (1937) showed that leaves maintained in the dark formed more asparagine than leaves maintained in the light, which synthesised glutamine preferentially. Further work by Yemm (1950) using excised barley leaves indicated that a succession of reactions followed protein breakdown, involving the production of high concentrations of glutamine, followed by asparagine and finally ammonia. A similar short-term accumulation of a high concentration of asparagine occurred following the incubation in the dark of leaves of Lolium temulentum (Thomas, 1978), oats (Malik, 1982) and wheat (Peeters & Van Laere, 1992). When detached shoots of P.sativum were incubated in the light for 72 h, there was little change in the concentrations of the major amino acids. However, if the shoots were incubated in the dark, the asparagine concentration (but not that of any other amino acid) increased from 3.17 to 24.0 μmol g−1 fresh weight (FW) during the 72-h period (Joy et al., 1983). Readers who are molecular geneticists will be pleased to learn that even Arabidopsis thaliana plants, when placed in the dark, accumulated high concentrations of asparagine after 6 days (Fig. 7) (Lin & Wu, 2004).
Very early studies showed that asparagine accumulated to high concentrations in the germinating seedlings of Vicia sativa and a range of Lupinus species. Table 1 shows data from the classic experiments of Schulze (1898) with Lupinus luteus, where it can be seen that the accumulation of soluble asparagine can account for more than 50% of the total nitrogen in the seedlings. Subsequent data showed that the asparagine was distributed throughout the seedlings, with values (expressed as % of DW) for cotyledons (3.73), plumules (1.41), stems (4.48) and roots (2.17). In stems, asparagine accounted for more than 66% of the total nitrogen. Later experiments by Prianischnikov (1922) showed that the addition of ammonia could stimulate the accumulation of asparagine even further in a range of germinating seedlings, provided a source of carbohydrate was available. Further work showed that as well as in legumes, Papaver somniferum, Pinus sylvestris and Tropaeolum majus also accumulated asparagine during germination, while Cucurbita pepo, Helianthus annuus and Linum usitatissimum accumulated asparagine and glutamine. The early work on the nitrogen metabolism of germinating seedlings has been discussed in detail by Chibnall (1939) and Mckee (1962).
Table 1. Nitrogen content of germinating Lupinus luteus seedlingsa
Ungerminated Dry Seeds
6 Days in Dark
15 Days in Dark
15 Days in Dark and 10 Days in Light
Data expressed as N content per 100 g of ungerminated seeds, taken from the original work of Schulze (1898), as reproduced by Chibnall (1939).
With the development of accurate and efficient automatic equipment for the analysis of amino acids and amides, the accumulation of asparagine during germination has been reinvestigated. Capdevila & Dure (1977) showed that in germinating cotton (Gossypium hirsutum) seedlings, the asparagine concentration increased from 280 nmol per cotyledon pair in the dry seed to 2885 nmol after 5 days in the dark and 5362 nmol after 5 days in the light. Slightly later, Elmore & King (1978), also investigating germinating cotton, indicated that the major site of asparagine accumulation was in the embryonic axis, where during the 5 day germination period, the soluble amino acid concentration increased from 4.7 to 723.7 μmol g−1 DW. In mung bean (Vigna radiata), Kern & Chrispeels (1978) demonstrated that the soluble amino acid concentration increased rapidly during the first 5 days of germination and that asparagine accounted for 50% of the nitrogen in the cotyledon exudates, despite the fact that the storage protein contained only 13.4% of aspartyl residues.
In a very thorough experiment on loblolly pine (Pinus taeda L.), King & Gifford (1997) initially incubated stratified seeds at 2°C for 35 days, which was followed by 12 days of germination at 30°C. As the radicle emerged, there was a pronounced change in the amino acid profile of the seedling (Table 2). Despite the fact that the storage protein of the megagametophyte contained high concentrations of arginine (50% of total N), it can be seen that after 12 days, the concentration of soluble asparagine in the seedling greatly exceeded the other major amino acids glutamine and arginine and accounted for 70% of the total soluble pool. Rozan et al. (2001) examined germinating seeds of a wide range of lentil (Lens) species and showed that asparagine was quantitatively by far the most important protein amino acid in seedlings of all species studied. The concentration of asparagine ranged from 18.96 to 62.24 mg g−1 DW following germination for 4 days, having increased dramatically from almost zero in the dry seed and represented 40–50% of the soluble amino acids. Even in Canavalia ensiformis, a seed known to accumulate high concentrations of the non-protein amino acid canavanine, asparagine was present as a higher percentage of the total amino acid pool, 7 days after germination (Camargos et al., 2004). However, it should be noted that asparagine does not always accumulate during seed germination (Lea & Joy, 1983), e.g. Cucurbita moschata (Chou & Splittstoesser, 1972), Zea mays (Limami et al., 2002) and coffee (Coffea arabica) (Shimizu & Mazzafera, 2000).
Table 2. Composition of the major soluble amino acids of loblolly pine (Pinus taeda) seedlings during germination [data from King & Gifford (1997) with permission]a
Dry seeds were initially imbibed at 2°C for 35 days and then incubated at 30°C for the number of days shown. Data expressed as nmol per seedling, taken from King & Gifford (1997). ND, not determined.
Stress and asparagine accumulation
Asparagine also accumulates under conditions of stress. In some cases, this may be a direct biological response to the stress conditions, for example, by contributing to the maintenance of osmotic pressure. However, it may also be an indirect result of the restriction of protein synthesis under stress conditions.
The amino acid contents of leaves of a salt-sensitive wheat line were compared with a salt-tolerant amphidiploid, following treatment with increasing concentrations of NaCl. Asparagine accumulated in the younger leaves but not in the older, while proline exhibited the reverse of this trend. In the young leaves, the concentration of asparagine increased from 0 to 120 μmol g−1 DW following treatment with up to 200 mM NaCl in a dose-dependent manner (Fig. 8) while that of proline only increased to 10 μmol g−1 DW (Colmer et al., 1995). In contrast, in barley the concentration of proline was fivefold higher than that of asparagine, following treatment with 450 mM NaCl, irrespective of the ages of the leaves (Garthwaite et al., 2005). In the source photosynthetic tissues of Coleus blumei, asparagine increased from 0.2 to 3.0 nmol g−1 FW in 10 days following treatment with 60 mM NaCl and 12 mM CaCl2, a concentration second only to that of arginine. However, in the sink non-photosynthetic leaf tissue of C.blumei, the concentration of asparagine increased from 1.0 to 9.0 nmol g−1 FW 10 days after the addition of saline, which was twice that of arginine (Gilbert et al., 1998).
With the exception of molybdenum, which is a constituent of both nitrogenase and nitrate reductase (Schwartz & Mendel, 2006), the growth of plants in a medium deficient in an individual mineral ion, particularly if there is a plentiful supply of nitrogen, can stimulate the accumulation of high concentrations of asparagine. A review of the early literature by Stewart & Larher (1980) indicated that deficiencies in potassium, sulphur, phosphorus and magnesium were well documented in a range of plants and that for tomato, deficiencies of micronutrients, in particular zinc, also stimulated large increases in asparagine concentration (Possingham, 1956).
Although phosphate deficiency is frequently followed by arginine accumulation (Rabe & Lovatt, 1986), again there is evidence that asparagine also increases at the same time. In soybean deprived of phosphorus for 20 days, considerable amounts of asparagine accumulated in the roots and stems, while arginine accumulated in the leaves (Rufty et al., 1993). Higher asparagine concentrations were detected in the root and in particular the shoot of young tobacco plants deprived of a phosphorus supply for 10 days (Rufty et al., 1990). Similarly, asparagine accumulated in the nodules and the roots of white clover subjected to decreasing concentrations of phosphorus (Almeida et al., 2000).
When plants are grown under sulphur-deficient conditions, there is a reduction in the synthesis of the amino acids cysteine and methionine and the antioxidant glutathione (Nikiforova et al., 2006). In maize seeds, such a deficiency in sulphur has been shown to give rise to an alteration in the proportion of types of protein formed and an increase in soluble asparagine to 50% of the total pool (Baudet et al., 1986). In maize suspension culture cells, the increase in asparagine accumulation following sulphur starvation was shown to be as a result of de novo synthesis and not protein hydrolysis (Amancio et al., 1997). Similar increases in soluble asparagine in the shoots of sulphur-deficient wheat varieties (Zhao et al., 1996) and Italian rye grass (Mortensen et al., 1993) have been observed.
Plants exposed to toxic metals have been shown to accumulate specific amino acids such as proline and histidine, which may have a beneficial function (Sharma & Dietz, 2006). However, there is evidence that asparagine can bind to cadmium, lead and zinc (Bottari & Festa, 1996). It was suggested that in the grass Deschampsia cespitosa, the asparagine which accumulated in the roots of ammonium-grown plants formed an intracellular complex with zinc and thereby decreased its toxicity (Smirnoff & Stewart, 1987). White et al. (1981) were able to demonstrate the presence of asparagine–copper complexes in the xylem of soybean and to a lesser extent in tomato.
Cadmium is a major pollutant worldwide and is detrimental to plant growth (Benavides et al., 2005; Gratão et al., 2005). For lettuce plants growing in axenic hydroponic culture, cadmium caused an increase in asparagine in both the roots and shoots (Costa & Morel, 1994). Similar data of cadmium-induced asparagine accumulation in roots and to a lesser extent in shoots were also obtained in lupins (Fig. 9) (Costa & Spitz, 1997). Cadmium induced an almost 10-fold increase in the asparagine concentration of tomato roots but only fourfold in the leaves. However, glutamine was still the major amino acid being transported in the xylem and phloem and was little affected by cadmium (Chaffei et al., 2004).
When detached tomato leaves were maintained under sterile conditions for 10 days, the asparagine concentration increased from 25.7 to 1150 nmol g−1 FW. However, if the leaves were infected with the bacteria Pseudomonas syringea, the asparagine concentration rose to 19 843 nmol g−1 FW, an 800-fold increase from when the leaves were still on the plant (Perez-Garcia et al., 1998). There was evidence that a novel form of cytosolic glutamine synthetase was synthesised in the infected leaf mesophyll cells, which was involved in the dramatic increase in the rate of asparagine accumulation.
Following the infection of cocoa (Theobroma cacao) with witches’ broom, the basidiomycete fungus Crinipellis perniciosa, visible signs of chlorotic leaves and stem swelling were observed after 21 days. Asparagine, which was already present in high concentrations in the leaves, increased to even higher levels after this time (Scarpari et al., 2005).
The enzymes of asparagine metabolism
Asparagine synthetase (AS)
The major route of asparagine synthesis involves the initial assimilation of ammonia to the amide position of glutamine (Lea & Miflin, 2003), followed by the transfer to form the amide position of asparagine (Ta et al., 1986; Rhodes et al., 1989). The enzyme AS (EC 18.104.22.168) catalyses the adenosine triphosphate (ATP)-dependent transfer of the amino group of glutamine to a molecule of aspartate to generate glutamate and asparagine:
It has also been proposed that the enzyme can use ammonia directly as a substrate (Oaks & Ross, 1984), particularly if the concentration is high, but the in vivo operation of this pathway has not been clearly demonstrated. Aspartate is synthesised by transamination of oxaloacetate and is also an important precursor of the essential amino acids, lysine, threonine, methionine and isoleucine (Azevedo et al., 1997, 2006; Ferreira et al., 2006).
Asparagine may also be formed following the detoxification of cyanide (Piotrowski & Volmer, 2006) and the transamination of 2-oxosuccinamic acid (Joy, 1988), but the fluxes through these pathways are not thought to be of any significant magnitude (Sieciechowicz et al., 1988c). In this article, we will concentrate on the enzyme AS.
Measuring the expected levels of activity of the enzyme AS in plant tissues has proved a difficult task. On many occasions, authors have reported very low or zero levels of activity, although the particular plant tissue has been shown to be synthesising asparagine at high rates (Kern & Chrispeels, 1978; Joy et al., 1983; Brears et al., 1993; Hurst & Clark 1993; Chevalier et al., 1996; Seebauer et al., 2004). This inability to measure AS activity may have been because of an inherent instability of the protein, the presence of inhibitors or the competing enzyme asparaginase. More recently, attempts have been made to improve the extraction assay of AS in crude plant extracts using a complex extraction buffer, containing EDTA, MgCl2, ATP, aspartate, glycerol, DTT, and β-mercaptoethanol, followed by detection of 14C-asparagine using high performance liquid chromatography (Romagni & Dayan, 2000).
The purification to homogeneity of AS from alfalfa root nodules has been reported (Shi et al., 1997), but no details of the methods used or the kinetic properties were provided. Gálvez-Valdivieso et al. (2005) were able to express both genes encoding AS isolated from Phaseolus vulgaris in Escherichia coli. The PVAS2-encoded protein was used to raise antibody that recognised both the P.vulgaris gene products with a molecular mass of 65 kDa. AS protein was detected in mature roots, senescing leaves and only very early in the development of the root nodules of P.vulgaris (Gálvez-Valdivieso et al., 2005). Again unfortunately, no attempt was made to study the properties of the enzyme protein.
The genes encoding AS
The breakthrough in our understanding of plant AS came with the isolation of two genes encoding the enzymes AS1 and AS2 from peas by Tsai & Coruzzi (1990, 1991). The two proteins of molecular mass 66.3 and 65.6 kDa are highly homologous with the human enzyme and have a purF glutamine-binding site at the NH2 terminus, consisting of a Met-Cys-Gly-Ile sequence, indicating that glutamine is the likely substrate in vivo (Richards & Schuster, 1998). Northern blot analysis indicated that expression of both genes is repressed by light in the leaves but that in the roots, AS2 is expressed constitutively and only AS1 is repressed by light. The repression of the genes encoding AS by light and stimulation in the dark agrees with the early work showing that asparagine accumulation was stimulated by darkness.
An AS complementary DNA (cDNA) clone was isolated from asparagus spears that encoded a 66.5 kDa protein that was 81% identical to the AS1 from pea (Davies & King, 1993). Expression of the gene was not detected in the asparagus spears immediately on harvest, but after 6 h, there was a rapid induction of messenger RNA (mRNA) synthesis along the spear, but expression was not affected by light (Davies et al., 1996; Eason et al., 1996). A similar induction of expression of AS was also shown in harvested broccoli florets (Downs & Somerfield, 1997). It was proposed that the induction of AS mRNA was stimulated by a rapid reduction in the soluble sugar content. Although this proposal was confirmed using callus cultures, it proved experimentally difficult in excised asparagus spears (Irving et al., 2001). Later analysis of the promoter of the asparagus gene identified a potential carbohydrate-responsive element at −410 to −401 bp relative to the translation initiation ATG, with sequence identity to a rice α-amylase carbohydrate-responsive element (Winichayakul et al., 2004a). Further studies confirmed that low carbohydrate but not darkness acted as the signal for the induction of the promoter of asparagus AS (Winichayakul et al., 2004b), probably through the involvement of hexokinase.
This simple story of carbohydrate regulation became somewhat complicated when it was found that there were three genes encoding AS in A.thaliana (Lam et al., 1994, 1998), which appeared to be regulated in totally different manners. Expression of ASN1 was stimulated when plants were placed in the dark and dramatically repressed following exposure to light for only 2 h, while sucrose could to some extent substitute for light. In contrast, the expression of ASN2 was induced during the same period and further stimulated for another 16 h in the light; again sucrose could substitute for light. Even more interestingly, the expression of ASN1 was stimulated by the amino acids asparagine, glutamine and glutamate, while ASN2 was repressed by the same amino acids. Expression of the ASN3 gene was not detected in any of the organs examined. Further studies in A.thaliana by Thum et al. (2003) indicated that light was able to override carbon in the regulation of ASN1, while carbon was able to override light as the major regulator of ASN2. There was also evidence that blue and red light had differential effects on the expression of the AS genes. Wong et al. (2004) went on to demonstrate in A.thaliana that the light induction of ASN2 is ammonium dependent. In addition to ammonium, stresses such as salinity and cold also increased ASN2 mRNA levels, and these stresses correlated with increases in internal ammonium ions.
Three distinct genes encoding AS have also been identified in sunflower (H.annuus) (Herrera-Rodríguez et al., 2002, 2004, 2006). HAS1 and HAS1.1 were shown to be light-repressed genes whose transcripts accumulated to high levels in darkness. Light regulated the genes by means of two different mechanisms, a direct one, via phytochrome, and an indirect one, stimulating photosynthetic CO2 assimilation and the production of carbon metabolites such as sucrose. The third AS gene of sunflower, HAS2, was regulated by light and carbon in an opposite manner to that of HAS1 and HAS1.1. HAS2 was more widely expressed and was stimulated by light and by sucrose. HAS1 and HAS1.1 expression were dependent on the presence of a nitrogen source, while HAS2 transcripts were still found in N-starved plants. High ammonium levels induced all three AS genes and partially reverted the sucrose repression of HAS1 and HAS1.1 (Herrera-Rodríguez et al., 2004).
To investigate the involvement of asparagine and AS genes in the main nitrogen mobilisation processes in sunflower, the expression of HAS1, HAS1.1 and HAS2 genes, as well as the synthesis of asparagine and other nitrogen and carbon metabolites, were studied during germination and natural senescence of cotyledons and leaves (Herrera-Rodríguez et al., 2006). HAS2 was expressed early in germination, and there was a correlation between AS transcript level and asparagine accumulation in the sunflower tissues. During leaf senescence, all three genes were expressed, during which time there was a reduction in sucrose content. Somewhat surprisingly HAS1 and HAS1.1 were not expressed during cotyledon senescence.
Although there is considerable variation between plants in the exact mechanisms involved in the regulation of the expression of AS, there is an overall consensus. The expression of one gene (often that which is most highly expressed) is induced by a reduction in soluble carbohydrate supply and in some cases darkness, while a second gene is more widely expressed but may be stimulated by carbohydrate and light. An increased supply of reduced nitrogen, either as ammonium or amino acids, induces expression of AS genes.
Analysis of the amino acid sequences of plant ASs shows that the proteins contain glutamine, aspartate and AMP binding sites and are related to the E.coli asparagine synthetase ASB glutamine-dependent enzymes (Cañas et al., 2006). Phylogenetic trees of the plant amino acid sequences have been constructed by a number of researchers and compared with those of the bacteria, fungi and animals (Shi et al., 1997; Osuna et al., 2001; Moller et al., 2003; Cañas et al., 2006). In the most recent study, the plant sequences were clustered in two main groups: (a) the sequences close to A.thaliana AS1 and (b) those grouped with A.thaliana AS2 and AS3. The legume sequences were located in the AS1 cluster, while the monocot sequences were in the AS2/3 group (Cañas et al., 2006).
There are two established pathways of asparagine catabolism in higher plants, and these have been considered in detail by Ireland & Joy (1981), Joy (1988) and Sieciechowicz et al. (1988c). Asparagine can be transaminated, particularly in leaves, to yield oxosuccinamic acid, which may then be reduced to hydroxysuccinamic acid and subsequently deamidated to yield malate. The major form of asparagine, glyoxylate aminotransferase, has been studied in detail in peas (Ireland & Joy, 1983a,b). It is likely that asparagine is metabolised through the above route as part of the photorespiratory nitrogen cycle (Murray et al., 1987; Keys, 2006) but that the majority of the nitrogen is continuously recycled and that there is little net catabolism of asparagine.
Asparaginase (EC 22.214.171.124) catalyses the hydrolysis of asparagine to yield aspartate and ammonia. The ammonia is subsequently reassimilated by glutamine synthetase:
The assay of asparaginase has also proved difficult in higher plants, with some plant sources providing extracts with high rates of activity and others low or zero; a full description of the early setbacks has been provided by Sieciechowicz et al. (1988c).
It was the detailed investigation by Atkins et al. (1975) that first gave an important indication of a reliable source of plant material that would be useful for studying asparagine metabolism. The investigators fed [14C]- and [15N]-labelled asparagine to the shoots of L.albus that were producing seeds inside pods. Most of the 15N in the endosperm fluid was recovered as ammonia, glutamine and alanine, while the 14C was not present in amino acids. As the seed developed, both 15N and 14C were found to be present in the amino acids of the seed storage proteins. Particularly high asparaginase activity was detected during the maturation development of the L.albus cotyledons (Atkins et al., 1975). Subsequently, a number of workers confirmed the presence of high activities of asparaginase in legume seeds during the maturation process (Lea et al., 1978; Murray & Kennedy, 1980; Chang & Farnden, 1981). An unusual twist to the story came when Sodek et al. (1980) described the presence of an asparaginase in both the testa and maturing cotyledons of peas that was totally dependent on the presence of potassium ions. Gomes & Sodek (1984) then went on to demonstrate that the asparaginase in developing soybean cotyledons was also K+ dependent. In soybean cotyledons cultured in vitro, the level of activity of the K+-dependent asparaginase was greatly reduced by the supply of glutamine (Tonin & Sodek, 1990).
A K+-dependent asparaginase has also been studied in some detail in pea leaves, where activity increases in the light and decreases in the dark (Sieciechowicz et al., 1985). The use of inhibitors of protein synthesis and proteases demonstrated that this diurnal variation in asparaginase activity was because of enzyme synthesis in the light and proteolytic degradation in the dark (Sieciechowicz et al., 1988a,b; Sieciechowicz & Ireland, 1989a). It was proposed that this type of regulation ensured that the enzyme is only functional in the light when there is sufficient ATP, and reducing power to fuel the glutamate synthase cycle (Sieciechowicz et al., 1988a,b).
The K+-independent asparaginases isolated from different lupin species were shown to have similar properties including Km values for asparagine (4–12 mM) and pH optima (8.0–8.5) (Lea et al., 1978; Chang & Farnden, 1981). It was originally suggested that the native enzyme from Lupinus polyphyllus seeds was a dimer of molecular mass of 71–72 kDa with subunits of 35–38 kDa (Lea et al., 1978; Sodek & Lea, 1993). Lough et al. (1992a) went on to show that when asparaginase was purified from Lupinus arboreus seeds, although the native molecular mass was 75 kDa, three polypeptides in the range 14–19 kDa were present following sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE). At the time, the reason for the multiple number of subunits was not clear. Sodek et al. (1980) reported a Km for asparagine for the K+-dependent asparaginase in pea as 3.2 mM for the cotyledon and 3.7 mM for the testa enzyme and a native molecular mass of 68 kDa. Similar Km values were determined for the K+-dependent enzyme in pea leaves but with a lower native molecular mass of 58 kDa (Sieciechowicz & Ireland, 1989b).
The genes encoding asparaginase
A cDNA clone encoding a K+-dependent asparaginase was isolated from L.arboreus. This encoded a 32.8 kDa protein, which appeared to be only expressed at a specific time during seed maturation coinciding with high enzyme activity. Somewhat surprisingly, the gene was not expressed in roots, which had also been shown to have high asparaginase activity (Chang & Farnden, 1981; Lough et al., 1992b). Dickson et al. (1992) isolated a genomic sequence encoding asparaginase from Lupinus angustifolius that contained four exons and three introns. The 5′-flanking region contained sequences associated with nodule-specific and seed-specific expression. A genomic clone encoding asparaginase was also isolated from A.thaliana, encoding a protein with a predicted molecular mass of 33 kDa (Casado et al., 1995).
The promoter of the asparaginase gene isolated by Dickson et al. (1992) was ligated to a β-glucuronidase (GUS) reporter gene and transformed into tobacco plants (Grant & Bevan, 1994). GUS activity was found mainly in the developing tissues of mature plants such as apical meristems, expanding leaves, inflorescences and seeds of tobacco. The chimaeric gene was also used to investigate transient expression in lupins. As might be expected from earlier enzyme measurements, transient GUS expression was detected in the developing pods, seed testas and cotyledons.
Our understanding of the molecular structure of plant asparaginases took a leap forward when Hejazi et al. (2002) were able to express the A.thaliana gene in E.coli. The purified asparaginase protein was shown to comprise peptides of approximately 35, 24 and 12 kDa, following SDS-PAGE. The authors proposed that the two smaller peptides were the result of proteolytic cleavage and that the native protein rather than being a dimer was in fact an (αβ)2 tetramer. Analysis of the substrate specificity of the recombinant A.thaliana protein showed that the enzyme could use a range of β-aspartyl peptides as substrates, with β-aspartyl-phenylalanine and β-aspartyl-alanine having Vmax values close to that of asparagine.
Borek et al. (2004) expressed a gene encoding the L.luteus K+-independent asparaginase (Borek et al., 1999) in E.coli. The recombinant native enzyme had a molecular mass of 75 kDa, but the translated peptide underwent an autoproteolytic cleavage leading to the formation of two subunits, 23 kDa (α subunit) and 14 kDa (β subunit), confirming the existence of the (αβ)2 tetramer. This cleavage gives rise to a N-terminal nucleophilic threonine residue on the β subunit. Phylogenetic analysis of N-terminal nucleophilic hydrolases indicated that the amino acid sequences of the plant asparaginases from A.thaliana, L.luteus, barley, rice and soybean fell in a group with bacterial enzymes that also had isoaspartyl peptidase activity. Although asparagine was a substrate for the recombinant L.luteus enzyme with a Km of 4.8 mM, the surprising result was that β-aspartyl-leucine was a substrate with more than four times the Vmax and a Km of only 0.14 mM (Borek et al., 2004).
Michalska et al. (2006) crystallised the K+-independent asparaginase from L.luteus and carried out a detailed analysis of the quaternary structure. The protein exhibited a αββα fold typical of N-terminal nucleophilic hydrolases. Each of the two active sites of the (αβ)2 heterotetrameric protein is located in a deep cleft between the β-sheets, near the nucleophilic threonine 193 residue, which is liberated in the autocatalytic event at the N-terminus of the β subunit. A comparison of the active sites of the L.luteus asparaginase and the E.coli EcAIII enzyme showed a high degree of conservation of the residues participating in substrate/product binding and of all other residues forming important hydrogen bonds within the catalytic pocket. Some evidence was provided as to how the active site could accept both asparagine and β-aspartyl peptides.
The availability of the complete sequence of the A.thaliana genome allowed Bruneau et al. (2006) to isolate a second gene encoding an asparaginase enzyme that was dependent on K+ for full activity. The K+-dependent enzyme had 55% identity with the K+-independent form, indicating that they belong to two evolutionarily distinct subfamilies of plant asparaginases, as revealed by phylogenetic analysis. However, the two enzyme proteins had remarkably similar structures, the K+-dependent enzyme having α subunits of 22.7 kDa and β subunits of 13.6 kDa. In addition, there were conserved autoproteolytic pentapeptide cleavage sites commencing with the catalytic threonine nucleophile, as determined by electrospray ionisation-mass spectrometry (ESI–MS) analysis. The K+-dependent enzyme in A.thaliana had a lower Km and much higher Vmax than the K+-independent form, indicating an 80-fold higher catalytic efficiency with asparagine. The K+-dependent enzyme was unable to use β-aspartyl dipeptides as substrates, demonstrating a clear difference between the enzyme and the K+-independent enzyme (Bruneau et al., 2006).
The steady-state mRNA levels of the two asparaginase genes in A.thaliana were determined by quantitative reverse transcriptase polymerase chain reaction (RT-PCR) in various tissues during development. As expected, the expression of both genes was associated with sink tissues and was highest in flowers, siliques, flower buds and leaves. The two genes showed largely overlapping patterns of developmental expression, but in all the tissues examined, the transcript levels of the K+-dependent enzyme were lower than those of the K+-independent enzyme. Microarray analysis showed that the K+-dependent enzyme was highly expressed in stamens and mature pollen of A.thaliana (Schmid et al., 2005).
Bruneau et al. (2006) suggested that as the spatial patterns of the expression of the two genes were largely overlapping, the two enzymes had redundant functions. As mutants and knockout lines are not currently available, it is not possible to test this hypothesis. However, the key question is why should plants have one form of an asparaginase which apparently has a greater activity and a higher affinity for isoaspartyl peptides? One possible reason is because of the frequently occurring conversion of asparagine to isoaspartyl residues in mature proteins. This is a dangerous modification, as it causes a structural change that may significantly alter the three-dimensional structure of the protein, leading to a change of activity, degradation or aggregation. Proteins with isoaspartyl residues can be degraded by proteolytic enzymes, but among the products, there will be β-aspartyl peptides containing N-terminal isoaspartyl residues which require specialised hydrolytic enzymes (Clarke, 2003; Shimizu et al., 2005). Borek et al. (2004) proposed that isoaspartyl peptidase activity could be particularly important in seeds that have to retain their ability to grow for a very long time. During the storage period, the seed proteins can undergo modification, and isoaspartyl peptidase activity is necessary to destroy the altered proteins and to allow only the healthy seeds to germinate.
Asparagine in crop plants
The accumulation of asparagine in the edible organs of crop plants is of particular interest in relation to acrylamide formation in food (discussed below). Current information indicates that both genetic and environmental factors may be important, with the latter being of particular concern in relation to the sourcing of low asparagine cereals and potatoes for food processing.
There are relatively few detailed studies of free amino acids in cereal seeds, with most emphasis being on the protein fractions which determine nutritional quality and functional properties. Further, many early studies often determined free amino acids as part of the ’non-protein nitrogen’ (NPN) fraction that would include other water-soluble, low molecular weight nitrogenous components including peptides. Thus, Byers et al. (1983) showed that the NPN extracted with three concentrations of NaCl solution accounted for 6.4–6.7% of the total grain N of wheat, and similar values of 5–6% were reported for three wild type (i.e. non-mutant) barley lines by Køie & Doll (1979). In maize, total free amino acids have been determined as 4.4% and 2.9% of the total grain nitrogen in two hybrids with normal grain texture (Sodek & Wilson, 1971). Even less data are available on the contents of asparagine in these fractions, but values of 3.31% of the total free amino acids have been reported for five inbred rye lines with low protein content (Dembinski & Bany, 1991) and 7.9% for a single inbred line of maize (Sodek & Wilson, 1971). Hence, it can be concluded that free amino acids generally account for about 5% or less of the total nitrogen in cereal grains, and asparagine accounts for a low proportion (certainly less than 10%) of this fraction. However, these values may be dramatically affected by both genetic and environmental factors.
Køie & Doll (1979) analysed 11 induced high-lysine mutants in the barley cultivars Bomi and Carlsberg II and showed that these had up to threefold increases in free amino acids (expressed as mg g−1) and up to twofold increases in NPN. However, amino acid analyses of these fractions were not reported. More extensive studies have been reported of maize mutants, showing substantial increases in free amino acids but that the magnitude of these varies between mutant lines and reports (Murphy & Dalby, 1971; Ma & Nelson, 1975; Misra et al., 1975; Arruda et al., 1978). One of the most detailed studies was reported by Sodek & Wilson (1971). They showed that free amino acids were increased from 4.4% to 19.7% of the total N in an opaque2 line in the R802 background, with other normal (WF9×M14) and opaque2 (R802×R75 opaque2) lines having 2.9% and 9.1% free amino acids, respectively. In contrast, the amount of free amino acids was lower in the R802 floury2 line (2.2%) than in the normal R802 background (4.4%). These authors also determined the amino acid compositions of free amino acid fractions from the R802 and R802 opaque2 lines, showing 7.9% and 7.2% asparagine, respectively. Similarly, Arruda et al. (1978) compared developing endosperms of a double mutant (sugary1/opaque2) line of maize with the normal background line. Total free amino acids and asparagine were determined as μmol per endosperm, and the ratios of these in the mutant : normal endosperms were similar, 1.7:1 for free amino acids and 1.85:1 for asparagine. Thus, it appears that there is little or no impact of the mutations on the proportion of asparagine in the fractions.
Impact of nutrition
Winkler & Schön (1980) grew barley plants in pots under glasshouse conditions with four nutrient regimes: 0.6 g N per pot, 1.2 g N per pot as single and split applications and 1.8 g N per pot as a split application (Table 3). Although the total free amino acids and total asparagine increased with grain nitrogen, the proportion of asparagine remained constant at about 15% of the fraction. Thus, increasing seed N did not disproportionally affect asparagine accumulation.
Table 3. Total grain N, free amino acids and free asparagine in barley grain grown with various levels of nitrogen [data from Winkler & Schön (1980), with permission]
Fertilisation (g N per pot)
0.6 + 0.6
1.2 + 0.6
Total grain N (%)
Total free amino acids (μmol per 100 mg flour)
Free asparagine (μmol per 100 mg flour)
Free asparagine (% total free amino acids)
However, subsequent work on barley by Shewry et al. (1983) indicated that the ratio of S:N was an important determinant of asparagine accumulation rather than N alone. They grew plants under sulphur-deficient conditions, which restricted their ability to synthesise the major hordein (prolamin) storage proteins that contain cysteine and methionine. Under these conditions, total hordein was decreased from 51.0% to 27.0% and from 46.7% to 26.9% of the total seed N in two cultivars, and NPN increased from 6.8% to 29.1% and from 7.3% to 32.6%, respectively. The amino acid composition of the NPN fraction was not determined, but analysis of total grain demonstrated increases in the proportion of aspartate + asparagine to 19.2% and 18.5% of the total amino acids from 5.7% and 5.3%, respectively. The authors suggested that this resulted from the accumulation of free asparagine that acted as an alternative nitrogen store when the ability to synthesise hordein was compromised.
When the study of Shewry et al. (1983) was carried out, the results were of little more than academic interest as sulphur deficiency was not an issue for cereal cultivation, except in some specific parts of the world outside Europe. This situation has now changed, with increasing areas of land in Western Europe becoming sulphur deficient as a result of decreased use of sulphur-containing fertilisers and reduced atmospheric deposition (Zhao et al., 1999). In fact, the Home Grown Cereals Authority currently estimates that 23% of the UK is at risk of sulphur deficiency for cereal cultivation (www.hgca.uk; verified 3/11/2006). Consequently, Muttucumaru et al. (2006) have carried out a detailed study of wheat grown in glasshouse trials and in a field trial on the sulphur-deficient site at Woburn (Bedfordshire, UK). This showed that similar high levels of asparagine accumulation (75 mmol kg−1 FW) occurred in grain grown in the field as in the highly sulphur-deficient grain grown in glasshouse experiments (48–153 mmol kg−1 FW) (Table 4). The implications of these levels for the formation of acrylamide during processing are discussed below.
Table 4. Free asparagine (mmol kg−1 FW) in grain samples from wheat grown in pots under glass with adequate and deficient levels of sulphur or in the field at the Rothamsted farm site at Woburn, Bedfordshire, UK, with the addition of sulphate fertiliser to give 0, 10 or 40 kg sulphur per hectare as indicated; the soil at this site has very poor nutrient retention and without the addition of fertiliser is severely sulphate deficienta
Baker et al. (2006) reported detailed metabolomic comparisons of a series of transgenic and control lines of wheat grown on two sites (Long Ashton, near Bristol, and Rothamsted, near London, UK) over 3 years. These included the analysis of free amino acids in samples grown in 2000 using gas chromatography-mass spectrometry (GC–MS) analysis. The transgenic lines had been engineered to express additional copies of gluten protein genes, and there were few differences between the amounts and compositions of free amino acids present in the control and transgenic lines. However, there were consistent differences between sites, with the levels of asparagine (Fig. 10) and several other free amino acids (aspartic acid, γ-aminobutyric acid, glutamine and glutamic acid) (results not shown) all being higher in the grain grown at Rothamsted than in that grown at Long Ashton.
These differences could not be accounted for based on the N and S contents of the grain, with the samples from Long Ashton having mean contents of 1498 mg S kg−1 and 2.45% N on a DW basis and those from Rothamsted 1710 mg S kg−1 and 2.73% N (unpublished results from the study). The N:S ratios for the two sites were therefore 16.36 for Long Ashton and 15.97 for Rothamsted, with Long Ashton being slightly more sulphur deficient. The two sites also differed in soil type and climatic conditions (temperature, precipitation), but the reason for the effect on asparagine accumulation has not been established.
Grain protein content
Dembinski & Bany (1991) determined the free amino acids in five inbred lines of rye with normal protein content (≈13% DW) and five lines with high protein content (≈20% DW). The high-protein lines contained about three times more free amino acids than the normal-protein lines (124.1 μmol g−1 DW compared with 45.4 μmol g−1 DW), while asparagine increased from 3.31% to 17.37% of the pool of free amino acids. The plants were grown with a modest level of nitrogen fertiliser (40 kg ha−1), but sulphur was not added. Hence, the increased accumulation in the high-protein lines could possibly be accounted for by the increased demand for nitrogen, resulting in sulphur deficiency.
Location within the grain
Fredriksson et al. (2004) determined the content of free asparagine in a range of milling fractions from grain of wheat and rye. The wholemeal flour from these species contained 0.51 and 0.48 g asparagine kg−1 DW for two batches of wheat and 1.07 g kg−1 DW for one sample of rye. In both cases, the sieved flour contained less asparagine (0.14, 0.17 for wheat; 0.53, 0.68 for rye) and the bran contained more (1.48 for wheat; 2.61, 3.18 for rye). However, the highest levels were present in wheat germ, with two fractions containing 4.88 and 4.99 g asparagine kg−1 DW.
In potatoes, asparagine has been reported to be the dominant free amino acid (33–59% as a percentage of the total free amino acids) (Eppendorfer & Bille, 1996). However, there are relatively few systematic studies that compare the effects of genotype or environment on free amino acids in tubers. In those experiments that have been conducted and that compared tubers from different cultivars grown at the same site, the asparagine contents vary within a comparatively narrow range, e.g. 1.54–1.93 mg g−1 FW in commercially important Belgian cultivars (De Wilde et al., 2005); 0.9–2.0 mg g−1 FW in two crisp processing and one table cultivar (Matsuura-Endo et al., 2006); 3.8–5.3 mg g−1 DW in three crisp processing cultivars (Olsson et al., 2004). This contrasts dramatically with those studies that compare tubers purchased from local markets. Vivanti et al. (2006) reported almost 50-fold variation (1.2–57.6 mg g−1 FW) in the asparagine content of tubers for 31 (9 Italian and 22 American) cultivars purchased at markets. The extent to which this variation in asparagine content is because of genotype or other environmental factors is unclear, although the data presented above suggest that it is unlikely to be because of storage conditions. Unfortunately, there are relatively few detailed studies on the effects of nutrition or other environmental factors on the asparagine content of tubers. Several authors have reported that tubers have increasing asparagine concentrations in response to raised nitrogen fertilisation (Hoff et al., 1977; Amrein et al., 2003; Silva & Simon, 2005). In a more recent study, De Wilde et al. (2006) analysed tubers from three different cultivars of potatoes grown with three nitrogen regimes. A re-analysis of their data (Fig. 11) demonstrates that both the free asparagine content and the total free amino acid content were strongly positively correlated with the N availability (r2 of 0.88 and 0.75, respectively), although the percentage contribution of asparagine to the total amino acid pool remained almost constant. Earlier work by Osaki et al. (1995) showed that the form of nitrogen supply, ammonia or nitrate did not affect the free asparagine content of tubers. There are few systematic studies describing how the free amino acid content responds to changes in other major nutrients (P, K and S) or to environmental factors; a rigorous systematic analysis of these responses is needed.
The relationship between asparagine content and acrylamide formation during processing
It only recently became apparent that the accumulation of asparagine in the harvested organs of crop plants had implications for food safety. A study published in 2002 by a research group led by Margareta Törnqvist at the University of Stockholm found that people had a significant intake of acrylamide from cooked foods (Tareke et al., 2002). Later that year, research at the University of Reading provided the first evidence that acrylamide can be generated from food components during heat treatment as a result of the Maillard reaction (Fig. 12), which occurs between amino acids and reducing sugars and that asparagine was the amino acid required for its formation (Table 6) (Mottram et al., 2002). Other work has confirmed these findings (Stadler et al., 2002; Becalski et al., 2003; Zyzak et al., 2003). The International Agency for Research on Cancer has classified acrylamide as ’probably carcinogenic to humans’ (IARC, 1994). Carcinogenicity to humans has not been demonstrated in epidemiological studies (Mucci et al., 2003), but carcinogenic action in rodents has been demonstrated. At high doses, acrylamide also has neurological and reproductive effects. Those foods with the highest levels of acrylamide are carbohydrate-rich foods that have been cooked at high temperatures, such as those achieved during frying or baking (Table 7). They include foods derived from wheat, maize, rye, other cereals and potato. Levels of acrylamide can exceed 1000 parts per billion (ppb); for comparison, the tolerance level for water set by the World Health Organization is 1 ppb. To date, no tolerance levels have been established for food, and, with the exception of Germany, where the authorities work with a de facto limit of 1000 ppb, no country has yet specified maximum levels of acrylamide permitted in food products. However, the European Commission Scientific Committee on Food has recommended that levels of acrylamide in food should be as low as can be reasonably achieved. Note that acrylamide is not found in boiled or unheated foods and is only formed in low concentrations in heated protein-rich foods, such as meat.
Table 6. Acrylamide produced in reactions at 175°C between glucose and amino acids [based on data in Mottram et al. (2002), with permission]
Acrylamide (mg mol−1 amino acid)
Table 7. Acrylamide concentrations reported in some cooked foods and food products [data from Friedman (2003)]
Acrylamide [μg kg−1 (ppb)]
Potato chips, crisps
Snacks, other than potato
The detection of acrylamide in foods led to much effort worldwide by the food industry and regulatory authorities to monitor acrylamide levels and to find means of reducing the levels and thereby any possible health risks. The formation of acrylamide has been shown to be dependent on time and temperature of cooking (Amrein et al., 2003; Jung et al., 2003; Rydberg et al., 2003; Surdyk et al., 2004; Taubert et al., 2004), and progress has been made in some sectors of the food industry to reduce acrylamide formation by changing processing methods. Potato snack manufacturers, for example, claim to have reduced acrylamide levels in their products by approximately 40%, although the evidence for this is anecdotal. They also benefited from the fact that potato varieties bred for crisp and chip production had been selected for low sugar content for many years before the acrylamide issue arose to ensure the correct colour after cooking. However, the Maillard reaction is also responsible for the characteristic colour and flavour of roast, baked and fried foods, and any general inhibition of the Maillard reaction to lower acrylamide levels is likely to affect colour and flavour quality. If sugars levels are too low, such as may be achieved in potato products by washing before cooking, then flavour and colour are affected.
The demand for fried, roasted and baked foods is unlikely to diminish, and improvements brought about by changes in processing methods will have a limit. Attention must turn to the raw material, and several studies have shown that the levels of precursors in the raw material are an important factor (Amrein et al., 2003; Biedermann-Brem et al., 2003; Grob et al., 2003; Haase et al., 2004). Acrylamide in food is therefore a plant and agricultural science as well as a food science issue, and plant scientists, breeders and farmers must be engaged in addressing it. Muttucumaru et al. (2006) demonstrated a link between one agronomic parameter, sulphate availability, and asparagine accumulation in wheat grain, while Fig. 10 shows clear effects of location (Baker et al., 2006). These results suggest strongly that plant- and agronomy-based studies could make a significant contribution in reducing the levels of acrylamide in processed foods by improving the raw material.
The data reported by Muttucumaru et al. (2006) can be used to show a clear relationship between wheat grain asparagine and acrylamide formation during processing (Fig. 13). Strategies for lowering acrylamide levels in wheat and almost certainly other cereal products therefore need to reduce asparagine while keeping sugar at a level sufficient to maintain flavour and colour quality. The link between asparagine levels and acrylamide formation in potato is not so clear; reports by Ohara-Takada et al. (2005) and Wicklund et al. (2006), for example, suggested that sugars not asparagine levels were limiting in potato. However, in our view, more data are required before definite conclusions can be drawn because varieties may differ; it would be inappropriate at this point to recommend to breeders that they should focus their efforts entirely on sugars. We note that crisp manufacturers already use very low sugar varieties and still have acrylamide levels in their products of several hundred ppb. In varieties such as these, asparagine content may be a more useful target, but there are insufficient data available at this time to be certain.
Asparagine clearly plays a central role in nitrogen storage and transport in plants, which is facilitated by its chemical properties. This includes accumulation in a range of tissues (sometimes transiently) and under stress conditions, including conditions where the plant is unable to support a normal level of protein synthesis. The latter may account in part for the wide range in asparagine concentrations in harvested cereals and potatoes, and the consequences of this account for the formation of acrylamide in foods. However, there are clearly also genotypic differences, and to unravel these and how they interact with environmental factors is an important topic for future research.
Efficient and reliable transformation systems are now available for all major crops (Howarth et al., 2005; Shewry & Jones, 2005; Meiyalaghan et al., 2006) including cereals and potatoes, which are of greatest concern to food processors in relation to acrylamide formation. Determination of the factors and mechanisms that determine asparagine accumulation may therefore ultimately be manipulated to give safer food products for consumers.
Rothamsted Research receives grant-aided support from the Biotechnology and Biological Sciences Research Council of the UK. We are indebted to the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq-Brazil) for funding collaboration between L. S. and P. J. L.